The techniques of experimental design, response surface models, and a multi-objective genetic algorithm were studied for optimizing submersible shapes. Six parameters which determined the shape of revolution were selected, and second-order polynomial response surface models of drag and surrounded volume were constructed using these samples and responses obtained by hydrodynamic calculations. A multi-objective genetic algorithm was studied to solve the problem of shape optimization with two objectives, which were minimum drag and maximum surround volume, and the Pareto optimal solutions were given in the end. The result shows that this method can improve efficiency and achieve optimal intention. The approximation of drag for shape revolution and the graphs of influence from submersible shapes on drag provide valuable reference information for later research.%针对潜器的艇型优化,研究了试验设计、响应面模型技术和多目标优化算法.选择了决定回转体艇型的6个参数,将阻力数值计算结果进行数据回归,建立阻力、体积与型值的二阶多项式响应面模型.然后将阻力最小与体积最大作为回转体艇型优化的两个目标,研究了多目标遗传算法,并给出了 Pareto 最优解集.结果表明,在艇型优化过程中采用近似模型,不但效率得到提高,优化的效果也能得到保证.由数值计算结果建立的回转体艇型阻力估算公式以及艇型变量对阻力的影响关系可以为后续设计带来一定的参考价值.
展开▼
